© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 5803
    Prev
    Next
    INDEX
    Explanations

    documentation, file, forms, browse

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron strongly activates on capitalized, multi‐letter words that look like proper names or titles (e.g. “London,” “Cape,” “Social,” “Most,” “canceling”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    רג
    0.77
    onasera
    0.77
     motoc
    0.75
    eigen
    0.73
    ){
    0.71
    Dm
    0.71
     fotografi
    0.71
    }".
    0.70
     conducteur
    0.70
    .'''
    0.69
    POSITIVE LOGITS
    स्वी
    0.73
    ेंच
    0.72
    га
    0.71
    ंच
    0.68
    ociety
    0.68
    മ്പ
    0.68
     Teixeira
    0.68
     clín
    0.67
    几十
    0.66
     चि
    0.65
    Activations Density 0.000%

    No Known Activations